Non-negative Matrix Factorization for Face Illumination Analysis
Xianchun Zou, W. Wang, Josef Kittler · Open Research (University of Surrey) · 2008
Changing illumination causes severe problems for face recognition in uncontrolled environments. It might be helpful for illumination invariant face recognition if information about the illumination can be recovered from the given face image. In this paper an illumination classification method based on Non-negative Matrix Factorization(NMF) is proposed. The traditional NMF approach together with its few variants are investigated to classify an unknown illumination to one of the illumination conditions present in the training set. Encouraging results have been achieved on CMU-PIE face database which contains faces under various illumination conditions.